Optimized Area Coverage in Disaster Response Utilizing Autonomous UAV Swarm Formations

TL;DR

The paper tackles efficient disaster response with UAV swarms facing energy and collision risks. It fuses local ESDF-based obstacle avoidance, graph-based formation maintenance, and a PC-TSP–driven POI traversal to maximize area coverage under time windows. Key contributions include per-UAV local ESDF mapping, a formation similarity metric for robust coordination, a multi-term trajectory optimization, and MILP-based POI sequencing with time constraints. Simulation results demonstrate collision-free navigation, preserved formation, and complete area coverage in cluttered environments, highlighting practical potential for real-world wildfire monitoring and disaster management.

Abstract

This paper presents a UAV swarm system designed to assist first responders in disaster scenarios like wildfires. By distributing sensors across multiple agents, the system extends flight duration and enhances data availability, reducing the risk of mission failure due to collisions. To mitigate this risk further, we introduce an autonomous navigation framework that utilizes a local Euclidean Signed Distance Field (ESDF) map for obstacle avoidance while maintaining swarm formation with minimal path deviation. Additionally, we incorporate a Traveling Salesman Problem (TSP) variant to optimize area coverage, prioritizing Points of Interest (POIs) based on preassigned values derived from environmental behavior and critical infrastructure. The proposed system is validated through simulations with varying swarm sizes, demonstrating its ability to maximize coverage while ensuring collision avoidance between UAVs and obstacles.

Figures (5)

• Figure 1: System Overview. User-defined waypoints are prioritized, and a global trajectory is generated using a TSP variant. The UAV swarm is configured based on agent capabilities to maintain an optimal formation.
• Figure 2: Software architecture of UAV agent. The information exchange between agents includes the modules of localization, perception, control, and trajectory generation.
• Figure 3: Snapshots of a 4-UAV swarm in a 3D formation (red lines), avoiding obstacles - presented with blue tiles and marked (1) and (2). The yellow arrow points to the direction of motion. The trajectory of the central UAV is re-planned (green line) based on the presence of the obstacles. The blue dotted line marks the path that has already been completed. (a) The swarm, still in formation, is approaching the obstacle. (b)-(e) The swarm formation brakes to prevent collision with the obstacle in its path. (f) The swarm returns to the initial formation after evading the obstacle.
• Figure 4: POIs sorted with the TSP algorithm, based on distance, prize collection, and time windows (TWs). (a) The initial list of POIs. (b) The optimal path with relaxed TWs and equal prize values (prize=10) for all POIs. The swarm can visit all POIs inside the given TW and collect all prizes. (c) The optimal path with strict TWs and equal prize values (prize=10) for all POIs. The swarm visits the maximum number of POIs in the available TWs to collect the maximum prize possible. (d) The optimal path with strict TWs and uneven prize values (prize=20 for Point 4, prize=5 for Point 5 and prize=10 for the rest). The swarm visits the POIs, attempting to collect the maximum prize possible in the given TWs.
• Figure 5: A squared-shaped formation (dark green lines) consisting of 5 UAVs traverses the global path (pink line) from the prioritized POIs, starting from a home position. Snapshots display swarm behavior in different timestamps. Trajectory replanning is able to provide obstacle-free trajectories (green line) while maintaining the initial formation. The total area coverage is highlighted in yellow, the obstacles are presented as blue tiles, and the path already covered by each agent is shown by a blue line. The green and white dots on the paths are the control points of the B-splines of the future and past paths accordingly. (a) The drone is in an area with sparse obstacles and maintains the formation. (b)-(d) The swarm moves along an area cluttered with obstacles and the formation changes according to the generated paths.